This application focuses on the large number of children and adults who have diseases affecting the skeleton that are genetic, but not heritable. These disorders cause significant morbidity. Gorham-Stout Disease (GSD), Generalized Lymphatic Anomaly (GLA), bone cysts and tumors exemplify these diseases. The co-investigators of this application have developed an approach for identifying genetic causes for these types of disorders by combining massively parallel sequencing of DNA or RNA - recovered from a patient's affected tissues and unaffected tissues - with computational algorithms that detect somatic mosaic mutations that are either solely in sequence from affected tissue or are enriched in sequence from affected tissue compared to unaffected tissue. This approach does not rely on prospectively obtaining freshly excised lesional tissue, since it works using frozen tissue and archived formalin fixed paraffin embedded pathologic samples. We validated our approach by employing it to discover that somatic mosaic activating mutations in PIK3CA are responsible for CLOVES syndrome. We now propose to extend our work to other genetic, non-hereditary conditions. We have ascertained several well-characterized patient cohorts affected with diseases such as GSD, GLA, and bone tumors, and we have been collecting affected and unaffected tissue samples from these individuals. Our sample collection currently includes lesional tissue from 6 patients with GLA, 6 patients with GSD, 20 patients with giant cell tumors, 38 patients with chondroblastoma, 8 patients with chondromyxoid fibroma. In Aim One of this proposal, we will prepare bar-coded DNA and cDNA libraries from affected tissue and unaffected tissue, perform massively parallel sequencing, and use computational algorithms that have already enabled us to identify causative mutations in other disorders. We expect to successfully identify causative mutations in many of these new disorders. Aim Two of our proposal builds upon our discovery that PIK3CA activating mutations cause patterning defects, malformations, and overgrowths when present in somatic mosaic form. We have a mouse in which we can conditionally activate the p.1047H>R mutation, which we have found in persons with CLOVES, KTS, FIL, and isolated lymphatic malformations, and we will soon have a mouse in which we can conditionally activate and then inactivate a different disease-causing Pik3ca mutation, p.420C>R. We will use these mice to understand the pathogeneses of malformations and overgrowths that occur in PIK3CA-associated disorders and to determine whether these problems can be prevented, delayed, or reversed. Completion of these aims will link genes to phenotypes in patients whose diseases were previously impenetrable to genetic analyses and, for patients with PIK3CA-associated phenotypes, determine how lesions form and whether PIK3CA inhibition is a viable treatment strategy.